Advanced search in
Research products
arrow_drop_down
Simple Search
Searching FieldsTerms
Any field
arrow_drop_down
includes
arrow_drop_down
Include:
Filters (3)

Filters

Clear All
3 Research products, page 1 of 1

  • BASNUM
  • Hal-Diderot
  • Digital Humanities and Cultural Heritage

Relevance
arrow_drop_down
  • Publication . Conference object . Preprint . 2020
    Open Access
    Authors: 
    Pedro Javier Ortiz Suárez; Laurent Romary; Benoît Sagot;

    doi: 10.18653/v1/2020.acl-main.156

    Publisher: Association for Computational Linguistics
    Country: France
    Project: ANR | PRAIRIE (ANR-19-P3IA-0001), ANR | BASNUM (ANR-18-CE38-0003)

    International audience; We use the multilingual OSCAR corpus, extracted from Common Crawl via language classification, filtering and cleaning, to train monolingual contextualized word embeddings (ELMo) for several mid-resource languages. We then compare the performance ...

    addAdd to  ORCID orcid 
  • Publication . Doctoral thesis . 2020
    English
    Authors: 
    Khemakhem, Mohamed;
    Publisher: HAL CCSD
    Country: France
    Project: ANR | BASNUM (ANR-18-CE38-0003), EC | PARTHENOS (654119)

    Dictionaries could be considered as the most comprehensive reservoir of human knowledge, which carry not only the lexical description of words in one or more languages, but also the common awareness of a certain communityabout every known piece of knowledge in a time fr...

  • Publication . Conference object . Preprint . 2020
    Open Access English
    Authors: 
    Louis Martin; Benjamin Muller; Pedro Javier Ortiz Suárez; Yoann Dupont; Laurent Romary; Éric Villemonte de la Clergerie; Djamé Seddah; Benoît Sagot;

    doi: 10.18653/v1/2020.acl-main.645

    Publisher: HAL CCSD
    Country: France
    Project: ANR | PRAIRIE (ANR-19-P3IA-0001), ANR | SoSweet (ANR-15-CE38-0011), ANR | BASNUM (ANR-18-CE38-0003), ANR | PARSITI (ANR-16-CE33-0021)

    Pretrained language models are now ubiquitous in Natural Language Processing. Despite their success, most available models have either been trained on English data or on the concatenation of data in multiple languages. This makes practical use of such models --in all la...

    addAdd to  ORCID orcid